For the purpose of detecting motor fault many diagnostic methods have been developed so far. These methods to identify the induction machine faults may involve several different types of fields of science and technology. The most popular methods are:                temperature measurements;        infrared recognition;        noise and vibration monitoring;        chemical analysis;        acoustic noise measurements;        motor current signature analysis.        
Among them, the most widely uses in industry are vibration monitoring and current signature analysis. All of the above method required sensors for performing measurements and subsequent analysis. By traditional measurements of current or vibration actually not the root of the problem is detected but the consequences of magnetic flux leakage which often can lead to misinterpretation of results. In other words, in terms of the condition monitoring of electric machine, such as electric motor and electric generator, the electric machine's condition inferred from the measurement of the current or vibration of the electric machine is less accurate than that from the measurement of the magnetic field leakage of the rotor.
Besides, typically to perform condition monitoring of electric machine it is necessary to install sensors on electric machine and use data collector to measure respective signals. Such approach is expensive. Sensors and data collectors might already be expensive however additional cost is related with installation of sensors on the electric machine if customer does not have his own sensors and data collectors.